Re: mutopia's shortcomings

[Gilles]

Hi.

On Fri, 24 Apr 2015 12:43:59 +0200, Federico Bruni wrote:
> 2015-04-21 1:07 GMT+02:00 Kieren MacMillan 
> <address@hidden>:
>
> > > Seems to me it has been quite successful in its goals of making 
> > sheet
> > music easily available for free, all works in the public domain or 
> > under
> > creative commons licenses, in (user-editable, user-improvable) 
> > LilyPond
> > format, pdf, and midi — all with volunteer labor.  Looks like the 
> > total is
> > over 1900 works now.
> >
> > Other than the “user-editable, user-improvable” issue, all of those 
> > things
> > are far better done by IMSLP. Put another way, looking at IMSLP 
> > (with
> > 310,000 scores) and Mutopia (with 1,900), the shine quickly comes 
> > off
> > Mutopia for anyone except the handful of hardcore DIY musicians who 
> > (e.g.,)
> > want to take a violin piece from Mutopia and make a guitar 
> > arrangement.
> You forgot the quality of sheets: a (really) digital PDF will always 
> look
> better than a scanned PDF.
> And if it's too old or don't like anything you can change it and get 
> what
> you want.
>
> This is the value of Mutopia and the reason why I strongly disagree 
> on the
> idea of merging it into IMSLP. In the past discussion on this topic 
> there
> were a couple of ideas on better integration between the two 
> projects.
>
>
> > I think it would be far better — and probably result in better
> > visibility/marketing for Lilypond — if Mutopia were merged into 
> > IMSLP.
> > (There appears to have been a thought in this direction at some 
> > point, but
> > not any more; cf.
> >
> > http://imslp.org/wiki/IMSLP:Community_Projects/Mutopia_score_archive).
> > Then, for important works, there would be the Lilypond source, 
> > side-by-side
> > with scans of existing editions. But it seems this was considered, 
> > and
> > rejected for exactly the reasons that Mutopia now flounders (cf.
> >
> > http://imslp.org/wiki/IMSLP_talk:Community_Projects/Mutopia_score_archive
> > ).

Whether merge or not depends on the ultimate purpose of the project.
IMSLP is mainly a repository of "printed" scores (final output) that
happens to provide source code for some scores, while Mutopia is 
primarily
a repository of musical "data" (LilyPond input data) that happens to 
provide
the final output.

I think that the original problem with Mutopia is that it did not 
position
itself as a "contents" database but only as a score "repository".
In the former case, one would have required that submitted contents 
follow
rules well beyond just being a LilyPond-compilable source.

If all works would follow the same standard layout, it would be much 
easier
to maintain, upgrade (the layout) and adapt to different users taste 
wrt to
the output (for example, changing the font should be doable with just
rerunning LilyPond with an appropriate command-line switch).

Even if not everyone will agree on "the" standard layout, I feel that 
it
is extremely important to define one, with the maximum flexibility.
The contents of all files that make up the complete layout does not 
have to
be easily comprehensible by everyone; I think that the indispensable
features are that
* it should be manageable automatically (i.e. changing the standard
  should not require manual intervention)
* the files requiring user input (i.e. music contents) should be 
completely
  separate from layout definitions

Of course, the devil is in the details.
And power users will complain in advance that they must tweak things 
(i.e.
mix layout with contents) to get their required level of esthetics.
Maybe that tweaked editions should not be in Mutopia's realm as a 
database[1]
Maybe that such finely tuned editions could be managed with a source 
control
system (keeping track of the differences with the "baseline contents").


Best regards,
Gilles

[1] Those editions could be available from there too, but would not be 
(so
    easily) upgradable.